On Fri, Nov 15, 2002 at 11:15:20PM -0500, Peter S Galbraith wrote:
> Branden Robinson <branden@debian.org> wrote:
>
> > On Fri, Nov 15, 2002 at 08:48:40PM -0500, Jim Penny wrote:
> > > On Fri, Nov 15, 2002 at 08:03:53PM -0500, Branden Robinson wrote:
> > > > It hasn't so far. We have added packages to non-free faster than we've
> > > > been getting rid of them.[1]
> > > Anthony> total main contrib non-free %main %contrib %non-free
> > > Anthony> bo 1188 980 31 115 82.5 2.6 9.7
> > > Anthony> hamm 1852 1524 101 227 82.3 5.5 12.3
> > > Anthony> slink 2664 2269 97 298 85.2 3.6 11.2
> > > Anthony> potato 4305 3889 123 293 90.3 2.9 6.8
> > > Anthony> woody 8766 8291 203 272 94.6 2.3 3.1
> > > Anthony> ---------------------------------------------------------------
> > > Anthony> sarge 10283 9734 257 292 94.7 2.5 2.8
> > > Anthony> sid 11168 10555 306 307 94.5 2.7 2.7
> > >
> > > Odd. I would say
> > > 1) it has been essentially unchanged since slink.
> >
> > Well, if one wants to redefine "grown" as "essentally unchanged", yes.
>
> Huh? He did say since slink...
So? non-free has grown since slink; the numbers are right there.
Even using Mr. Penny's margins of error, "essentially unchanged" clearly
is not a *decline*, which is what opponents of the GR two years ago, and
opponents today, are telling us would happen. It is "inevitable", some
of them posit, that Free Software will produce alternatives to non-free,
and non-free will dwindle to zero.
The figures flatly contradict that hypothesis.
> > > 2) it has slowly been trending down. (with a bubble up in sid)
> >
> > 115 < 227 < 298 < 307
>
> 298 > 293 > 272
For the slink, potato, and woody *releases*, respectively. See below:
> > Packages that were unreleasable presumably weren't counted in the
> > distribution release totals. That the number of non-free packages in
> > unstable is larger than the number of non-free packages in any release
> > is instructive.
>
> For woody, it might mean lots of them weren't built for all releasable
> archs. Hardly an indicative sign since non-free is not autobuilt.
Isn't what matters the source package?
These numbers point up two problems with opposition to the GR:
1) What good are non-free packages doing our users if an ever-increasing
proportion of them are only available in unstable? If in 10 years we
have 700 non-free source packages in the archive, but "release" only a
few, what improvement have we really scored over letting some other
project handle non-free Debian packages?
2) The "bubble up in sid" indicates people's growing interest in
*maintaining* non-free packages, if nothing else. That trend is clearly
increasing. So where's the level-off? Where's the "inevitable
decline"? It looks like demand for non-free packages in *increasing*,
not decreasing, at least among Debian Developers, even if the users are
getting frozen out by autobuilder problems and releasability issues
("just run unstable", I guess they're told). If that's the case, why
aren't we amending the Social Contract to embrace non-free, instead of
relegating more and more software to second-class status?
In short, the status quo seems to be serving no one except people who
don't really care *what* the Social Contract says, which is a shame
given that all but the eldest Debian Developers vowed to abide by it
when we joined the Project.
If non-free is worth fighting for in Debian, why isn't it worth full
citizenship?
If non-free isn't worth fighting for, why not let it become some other
project's pride and joy, and not Debian's red-headed stepchild?
--
G. Branden Robinson | Don't use nuclear weapons to
Debian GNU/Linux | troubleshoot faults.
branden@debian.org | -- US Air Force Instruction 91-111
http://people.debian.org/~branden/ |
Attachment:
pgplfykc4KWpr.pgp
Description: PGP signature